Hi When I submit jobs in Hadoop how do the physical class files get distributed to the nodes on which the Map/Reduce jobs run ? Is some kind of dynamic class loading used or are the jar files copied to the machines where they are needed ?
Thanks Avinash
